The "transmitter hypothesis of chemoreception" postulates that the sensory discharge frequency of carotid nerve" fibers is governed by neurotransmitters released from glomus cells (GCs) in response to "natural stimuli (hypoxia, hypercapnia and acidity). Such transmitters cross the synaptic cleft between GCs and carotid nerve endings (NEs) to depolarize and excite the terminals. The main problem with this hypothesis is that specific synaptic blockers (while blocking the effects of exogenously applied to transmitters) do not block, only depress the sensory discharge elicited by natural stimuli. Thus, factors other than transmitter release should be involved in chemotransduction. We propose that GCs and NEs are coupled both by electrical and by chemical synapses as occurs in the avian ciliary ganglion and in several central nervous system synapses. Supporting this view is the recent study by Kondo & Iwasa showing that gap functions unite NEs to GCs. According to our hypothesis, most GC-NE junctions would be closed or non- functional at rest. During natural stimulation, the gap junctions would open, allowing ions and molecules to flow from cells to nerves. This hypothetical situation is the opposite of what occurs between apposed glomus cells (GC-GC channels) where natural stimuli mostly close intercellular channels. Specific transmitter antagonist that depress or eliminate activation of certain chemical synapses would not affect GC-NE gap junctions. Thus, elimination of chemical transmission still leaves intact GC-NE electrical transmission to signal changes in pO2 and pCO2. This idea will be tested in the rat carotid body with physiological and morphological experiments. We will measure macroscopic junctional conductance and intercellular channel activity between GCs and NES at rest and during stimulation using twin current and voltage clamping of GCs and NEs. Light and electron microscopy will establish whether transfer of markers occurs between these coupled structures during basal and stimulated conditions. Immunocytochemistry and freeze substitution ultrastructural techniques will determine the incidence of gap junctions between NEs and apposed GCs. This dual approach should establish the extent of interaction and physiological significance of electrical GC-NE gap junction synapses during carotid body activation.